Architecture search concepts in Computer Vision

Introduction

Architecture search helps find the best design for a machine learning model automatically. It saves time and effort compared to guessing the model structure.

When you want to build a new image recognition model but don't know the best layer setup.

When you want to improve an existing model's accuracy without manual trial and error.

When you have many possible model designs and want to find the fastest one that works well.

When you want to automate model design to save time for other tasks.

When you want to explore new model ideas that humans might not think of.

Syntax

Computer Vision

Define a search space of possible model parts
Choose a search strategy (e.g., random, evolutionary, reinforcement learning)
Run the search to test different models
Select the best model found

The search space is like a menu of model options to try.

The search strategy decides how to pick which models to test next.

Examples

This tries random models with 2 to 5 layers and picks the best one.

Computer Vision

Search space: number of layers = 2 to 5
Search strategy: random sampling
Run search for 10 models
Pick model with highest accuracy

This uses evolution ideas to improve models over time.

Computer Vision

Search space: convolution filter sizes = 3x3 or 5x5
Search strategy: evolutionary algorithm
Run search for 20 generations
Select model with best validation score

Sample Model

This code tries 10 random models with different layers and filter sizes. It prints each model's accuracy and shows the best one found.

Computer Vision

import random

# Define search space options
layer_options = [2, 3, 4]
filter_sizes = [3, 5]

# Dummy function to simulate model accuracy
# Higher layers and bigger filters give better accuracy here

def evaluate_model(layers, filter_size):
    base_accuracy = 0.7
    accuracy = base_accuracy + 0.05 * (layers - 2) + 0.03 * (filter_size - 3)
    noise = random.uniform(-0.01, 0.01)
    return accuracy + noise

# Simple random search
best_model = None
best_accuracy = 0

for _ in range(10):
    layers = random.choice(layer_options)
    filter_size = random.choice(filter_sizes)
    acc = evaluate_model(layers, filter_size)
    print(f"Tested model with {layers} layers and {filter_size}x{filter_size} filters: accuracy={acc:.3f}")
    if acc > best_accuracy:
        best_accuracy = acc
        best_model = (layers, filter_size)

print(f"\nBest model found: {best_model[0]} layers, {best_model[1]}x{best_model[1]} filters with accuracy {best_accuracy:.3f}")

Important Notes

Architecture search can take a lot of time if the search space is big.

Using a smart search strategy helps find good models faster.

Always test the final model on new data to check real performance.

Summary

Architecture search finds the best model design automatically.

It tries different model setups from a defined search space.

Choosing a good search strategy speeds up finding a good model.

Practice

1. What is the main goal of architecture search in computer vision models?

easy

A. To collect more training data

B. To manually tune model parameters

C. To automatically find the best model design

D. To reduce image resolution

Solution

1. Step 1: Understand architecture search purpose

   Architecture search aims to find the best model design automatically without manual trial and error.

2. Step 2: Compare options

   Options B, C, and D do not describe architecture search goals. Only To automatically find the best model design matches the goal.

3. Final Answer:

   To automatically find the best model design -> Option C

4. Quick Check:

   Architecture search = automatic best design [OK]

Hint: Architecture search = automatic model design finder [OK]

Common Mistakes:

• Confusing architecture search with data collection
• Thinking it manually tunes parameters
• Mixing it with image preprocessing

2. Which of the following is a correct way to describe a search space in architecture search?

easy

A. A set of possible model designs to explore

B. The training dataset used for the model

C. The final accuracy metric after training

D. The hardware used to run the model

Solution

1. Step 1: Define search space

   Search space is the collection of all possible model designs or configurations that the search will try.

2. Step 2: Eliminate incorrect options

   Options B, C, and D relate to data, metrics, or hardware, not the search space itself.

3. Final Answer:

   A set of possible model designs to explore -> Option A

4. Quick Check:

   Search space = possible designs [OK]

Hint: Search space = all model options to try [OK]

Common Mistakes:

• Confusing search space with dataset
• Thinking search space is a metric
• Mixing search space with hardware details

3. Consider this pseudocode for architecture search:

for model in search_space:
    accuracy = train_and_evaluate(model)
    if accuracy > best_accuracy:
        best_model = model
        best_accuracy = accuracy
print(best_accuracy)

What does this code output?

medium

A. The list of all models tested

B. The accuracy of the best model found

C. The training loss of the last model

D. The total number of models in search_space

Solution

1. Step 1: Analyze the loop

   The loop trains and evaluates each model, updating best_accuracy if current accuracy is higher.

2. Step 2: Understand the print statement

   After checking all models, it prints the highest accuracy found among them.

3. Final Answer:

   The accuracy of the best model found -> Option B

4. Quick Check:

   Prints best accuracy = highest accuracy [OK]

Hint: Code prints highest accuracy found during search [OK]

Common Mistakes:

• Thinking it prints number of models
• Confusing accuracy with loss
• Assuming it prints all models

4. The following code snippet is intended to find the best model architecture, but it has a bug:

best_accuracy = 0
for model in search_space:
    accuracy = train_and_evaluate(model)
    if accuracy < best_accuracy:
        best_model = model
        best_accuracy = accuracy
print(best_accuracy)

What is the bug?

medium

A. best_accuracy should start at 1 instead of 0

B. train_and_evaluate should return loss, not accuracy

C. The print statement should print best_model, not best_accuracy

D. The comparison operator should be > instead of <

Solution

1. Step 1: Understand the goal

   The goal is to find the model with the highest accuracy, so we want to update when accuracy is greater than best_accuracy.

2. Step 2: Identify the bug

   The code uses accuracy < best_accuracy, which updates for worse accuracy, so it should be accuracy > best_accuracy.

3. Final Answer:

   The comparison operator should be > instead of < -> Option D

4. Quick Check:

   Use > to find best accuracy [OK]

Hint: Best accuracy means use >, not < in comparison [OK]

Common Mistakes:

• Starting best_accuracy at wrong value
• Printing wrong variable
• Confusing accuracy with loss

5. You want to speed up architecture search by reducing the search space size. Which strategy is best?

hard

A. Limit model depth and number of layers to a smaller range

B. Increase the number of training epochs for each model

C. Use a slower but more accurate optimizer

D. Train all models on the full dataset without sampling

Solution

1. Step 1: Understand search space impact

   Reducing search space size means limiting the number of possible model designs to try.

2. Step 2: Evaluate options

   Limit model depth and number of layers to a smaller range reduces model complexity range, shrinking search space. Options A, B, and D increase training time or data size, slowing search.

3. Final Answer:

   Limit model depth and number of layers to a smaller range -> Option A

4. Quick Check:

   Smaller search space = fewer model options [OK]

Hint: Shrink search space by limiting model complexity [OK]

Common Mistakes:

• Thinking more training epochs speed up search
• Choosing slower optimizers to improve speed
• Using full dataset always speeds search